The SKA Molonglo Prototype
(SKAMP) Project
Molonglo 40th Anniversary, November 2005
SKA Molonglo Prototype Project
(SKAMP)
A new low-frequency spectral line instrument.
Funded by the ARC, the Science Foundation and the
Major National Research Facilities Program.
Project Goal: complete signal pathway – 2:1 dual
polarisation line feed; room temperature electronics;
wideband digital signal processing; FX correlator.
Features: wide field of view, imaging, polarisation,
spectral line capability, RFI mitigation (adaptive noise
cancellation).
Strategy: parallel 3-stage re-development of MOST
Science & technology prototyping for the Square
Kilometre Array (SKA) – 1% collecting area, wide-field
imaging.
What is the SKA?
Next generation radio
telescope – 100 times
improvement in many
parameters.
Global collaboration.
Proposed Australian site
in Mileura, WA.
Operational 2020.
6 Key science projects.
SKAMP Team
University of Sydney CSIRO – ATNF & ICT
Anne Green Michael Kesteven
Duncan Campbell-Wilson Tony Sweetnam
Adrian Blake John Bunton
Ludi de Souza Frank Briggs
Tim Adams John Kot
Martin Leung
Sergey Vinogradov Argus Technologies
Daniel Mitchell Bevan Jones
Elaine Sadler Martin Owen
3 site Technical Officers Peter Liversidge
Current Parameters for MOST
Single frequency - 843 MHz continuum
3 MHz bandwidth, RHC polarisation
43" spatial resolution
18,000 sq metres collecting area
Tsys 55K
Field of view: >5 square degrees
Sensitivity (7 position switching): 0.8 mJy
Sensitivity for full 12 hr: 0.3 mJy
Dynamic range: ~200:1
SKAMP 1 (2004 – 2005)
Continuum correlator: 96 station, 4.4 MHz
bandwidth, 843 MHz central frequency – > 4000
independent baselines, data rate 1sec
Sensitivity 0.8 mJy (12 hrs for complete
synthesis; 7-position switching to gain wide field
of view)
Continuous uv-coverage – correlation of inter-
arm & between-arm stations to give good image
fidelity
Programmable logic chips - FPGAs
Tim Adams & John Bunton
Continuum correlator
• 96 independent stations: 88
telescope bays + 2 reference
antennas
•Signal pathway complete -
commissioning at site
• Drift scan on sun – first light
SKAMP 2 (2005 – 2006)
Spectral-line capability: 830 - 860 MHz with 2048
channels via FX correlator.
Existing front end retained – 96 stations; full correlation
of all stations is highly redundant
Unchanged Tsys and angular resolution
Optic fibre distribution network designed – trenching and
conduit completed
Field of view roughly 4 square degrees
Sensitivity for 12 hrs observation: 0.15 mJy
Confusion limit of 0.12 mJy for 43" resolution
Spectral line measurements not confusion limited
Wide-band uncooled
Low Noise Amplifiers
Prototype 300-1000 MHz HEMT based LNA (Ralph Davison)
30 50
• ~20K noise temperature
• Ambient temperature Gain
Noise Temperature (K)
operation 40
• Possible extension to
operate 300-1400 MHz Gain (dB) 30
15
• Design simplified if
higher input impedance Noise Temperature 20
tolerated (50Ω input
impedance design now) 10
• Mass production (8000
units) requires simple 0 0
300 650 1000
assembly design Frequency (MHz)
Molonglo segmented parabola design
gives good performance to ~2 GHz
Piecewise linear fit to parabola shape
3 x focus
2
y (m)
Flat mesh tied
on supports at 1
points shown
0
-6 -4 -2 0 2 4 6
x (m)
• Mesh supported at 0.6 m (2 ft) intervals in x direction.
• Each section gives the same error for a linear fit to a parabola.
• 0.1 dB loss at 1420 MHz.
• f/D = 0.25
Simulations to test surface approximation
S. Vinogradov
SKAMP 3 (2006 – 2007)
Dual polarisation feed module – under range test. Next
stage to mount on Rapid Prototype Telescope (RPT).
Baseline ripples to be measured.
First feed prototype 700 – 1100 MHz. Instantaneous
bandwidth 100 MHz.
Once prototype approved, construct feeds for complete
RPT.
Stage 1 RF beamformer – switched delay lines, design set
by maximum frequency, ~3l length, 100 phase step gives
sufficient accuracy. Stage 2 beamforming also in feedline.
New mesh will reduce leakage to give Tsys of 40K.
12 hr sensitivity at 843 MHz ~0.1 mJy. Confused!
Polarisation not confusion limited (assume 5% mean
source value).
Wideband feed prototype module
1. 8-element module, 1.4 m length
2. Wide-band dipoles – no moving
parts
3. Polarisation axes oriented along &
across axis of feed – better
performance than dual-slant feeds
4. Range tested for 700 -1000 MHz
Leung
Beam & radiation patterns
1. Beam pattern – first sidelobe -
12dB; cross polarisation -30dB at
meridian, worst at high scan
angle, up to -12dB
2. Scanning gain curve – flat to ±45°;
cross polarisation -25dB or better
3. Transverse illumination pattern –
HPBW 80°; cross polarisation
worst at high scan angle, about -
Measured element patterns showing gain variation with scan angle (Transverse pol)
0
15dB
Measured element patterns showing gain variation with scan angle (Longitudinal pol)
0
-5 -5
-10
Figures show beam patterns
Normalised E-field (dB)
-10
Normalised E-field (dB)
-15 -15
and scanning gain variation for
-20 -20
the two polarisations,
-25 -25
transverse and longitudinal
-30 -30
-80 -60 -40 -20 0 20 40 60 80 -80 -60 -40 -20 0 20 40 60 80
Scan angle (deg) Scan angle (deg)
Rapid Prototype Telescope (RPT)
Visit by South African team
•Double mesh trial – reduce leakage
• Predict improve Tsys to 40K
•Construction of a 17m bay to test
feeds in realistic environment
RFI at Molonglo 200-1500 MHz
(Measured 25 June 2001)
Measured Power (dBm)
-75 VHF UHF
TV TV
-85
-95 GSM
-105
-115
0 500 1000 1500
Frequency (MHz)
RFI measurements in the field
at Dept of Defence HQ site
Campbell-Wilson, Briggs, Mitchell
Dual feed system for 6-m Reference
Antenna for adaptive noise cancellation
A further extension: uv-coverage with
additional stations on NS baselines??
•Good image fidelity in 6 hours
•Small reduction in sensitivity
•Double survey speed
•Model for 5 additional stations
(Bunton 2005)
Key science goals
1. Blind survey of HI absorption in high redshift
galaxies – initially z~0.7, extend later. Test of
mass-assembly of galaxies predictions from
CDM scenario.
2. HI in emission – measure mass function
directly. Redshift range z = 0.17 – 0.3.
Challenging.
3. Cosmic magnetism studies – measure
diffuse Galactic polarisation and a RM grid
from many extragalactic sightlines.
4. High redshift galaxies found as USS sources.
Molonglo continuum confusion
(10 beams/source) at δ = -60°
10
beam size:
112” x 112” csc|d|
Flux Density (mJy)
1
Rengelink et al 1997 beam size:
WENSS 325 MHz 43” x 43” csc|d|
beam size:
0.1 Bock et al 1999 26” x 26” csc|d|
SUMSS 843 MHz
Wall 1994
1420 MHz
0.01
0 500 1000 1500
Frequency (MHz)
High image fidelity results from good
uv-coverage
High-dynamic range continuum
imaging
StageI correlator will
allow self-calibration
strategies for MOST
Current MOST imaging
dynamic range is 100-
200:1 (similar to intrinsic
dynamic range of VLA)
(MGPS Green et. al.)
Self-calibration on VLA
enables imaging dynamic
ranges of more than 105:1
Current dynamic range of MOST limits imaging
of faint sources, such as filaments of
supernova remnants, near bright sources like
the Galactic Centre.
1. Blind HI-absorption survey
• New spectral line capability
• Measurements of HI
absorption at z ~ 0.75 that
capitalise on the large collecting (Lane 2000)
area of MOST
Stage 2: enables ΩHI measurements
at z ~ 0.75, where existing results
• >10,000 sightlines to search are not well constrained
for HI absorption – expect to
detect ~50 sources in limited
redshift range in 2400 sq deg
• Few detections – eg Darling et
al. (2004) of galaxy z=0.78 in
front of z=1.992 quasar. (Lane and Briggs 2001)
When & how is HI assembled into
galaxies? (Baugh et al 2004)
Data-free zone
2. High-redshift HI emission in
galaxies
11 HIPASS (500s) (12 h)
log10 Mlim (HI) (M⊙)
10 Molonglo
(10x12 h)
Typical
9 bright spiral
8
0 0.05 0.1 0.15 0.2 0.25 0.3
HI in the nearby Circinus
Redshift z galaxy (Jones et al. 1999)
The Molonglo telescope will reach HI mass limits typical of bright spiral
galaxies at z=0.2 (lookback time ~3 Gyr), allowing a direct measurement
of evolution in the HI mass function. Challenging project.
3. Cosmic Magnetism
Magnetism is crucial for :
– cloud collapse / star formation
– stellar activity / stellar outflows
– ISM turbulence / gas motions
Proplyd in Orion MHD turbulence
– supernova remnants
– stability of galactic disks
– acceleration / propagation /
confinement of cosmic rays
– heating in galaxy clusters
– AGNs / Jets SN 1006 Merger in gal. cluster
Magnetism is one of the fundamental forces in Nature,
but its role and origin is largely unknown !
Rotation Measure Grid
• Probes magnetic fields
in galaxies, the Milky
Way & clusters
• Rotation measure grid of
background sources and
polarisation of the diffuse
Galactic field
300 RMs through the LMC (Gaensler et al 2004)
4. High-redshift radio galaxies from
spectral studies, if lower frequency
range implemented
Radio spectral index measurements over
the range 300 –1400 MHz are an efficient
way of selecting high-redshift (z>3) radio
galaxies (e.g. de Breuck et al. 2000, 2004).
Radio galaxy TN0924-2201 at z=5.19
(van Breugel et al. 1999)
Summary of SKAMP Project status
96-station continuum correlator being commissioned –
first light. (SKAMP 1)
Optic fibre network conduit laid, fibre on order; spectral-
line correlator designed and being built, calibration &
image processing software being planned. (SKAMP 2)
8-element module of prototype feed under test; RPT
nearing mechanical completion; RF beamformers in
design. (SKAMP 3)
Simulated performance - sensitivity 0.12 mJy for 12 hour
observation – for 43” resolution, data are confusion
limited for continuum images but not for spectroscopy or
polarimetry
A new lease on life for a mature instrument
First Fringes – single baseline &
interim correlator
Team:
Green
Campbell-
Wilson
Kesteven
Bunton
Adams
*Leung
Blake
*Chippendale
Vinogradov
*Mitchell
Briggs
Sweetnam
Sadler